Molecular models of the 4th domain of CSF-1R have the site of oncogenic activation buried in the hydrophobic core of an immunoglobulin like domain. Unfolding of this domain caused by mutations introducing polar amino acids would lead to inter-molecular aggregation and activation of the CSF-1R kinase. Proposed course: Activating mutations of the CSF-1 receptor have been localized to the fourth extracellular domain. r S Molecular sequence alignment showed that the fourth domain of the CSF-1 receptor is most similar to immunoglobulin like domains. Using second heavy chain constant domain and the first light chain variable domain of the immunoglobulin structure 3HFM as a reference, molecular models for the fourth domain of CSF-mi 1R were built. These models were created by placing homologous atoms at the corresponding of sites of the reference structure and using the CHARMM empirical energy modelling program to place the remaining atoms and to relax the resulting structure to avoid chemically unreasonable bonds and non-bonded contacts. Molecular dynamics simulations were performed on models for the wild type (leucine) and lysine mutant (wildtype phenotype) as well as the activating mutations glutamic acid, serine. In all cases, simulations were run with the protein model surrounded by a ball of water (TIPS3 model). Variation in the exposed surface area and backbone hydrogen bonding structure were monitored. Although considerable variation between models based on the heavy and light chain domains was observed, the wildtype and wildtype-like mutant, lysine, appeared to be more stable than activated ms mutations.